Incentivizing Participation in an Internet-based Survey

How to cite

Smyk, Magdalena. 2023. "Incentivizing Participation in an Internet-based Survey." AEA RCT Registry. AEARCTR-0010760. https://doi.org/10.1257/rct.10760-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Magdalena Smyk*

* Primary investigator

Registry ID
AEARCTR-0010760
DOI
https://doi.org/10.1257/rct.10760-1.0
First registered
Last updated
Status
Completed
Trial dates
2020-09-26 – 2020-10-06
Countries
Poland ()
Registry record
http://www.socialscienceregistry.org/trials/10760

Abstract

Internet-based surveys provide data to social research especially when administrative sources are difficult to obtain or does not exist. In this context it is valuable to study effective ways to attract survey participants to enter and stay panel online survey. In this study, we experimentally test whether particular dimensions of the incentive to re-enter existing Internet-based panel survey conducted during pandemic lock-down in Poland can keep panel subjects in the survey. We find that providing a reward ex ante and lottery with a high award (but the same expected value as certain prize) increase response rate among survey participants. The long-lasting effect is observed only with an ex ante payment.

Experimental design

The experimental treatment was either financial or non-financial.
Financial treatment differ by: timing of the reward (ex-ante versus ex-post), expected value of the reward (two levels), probability of the reward (lottery versus certain), high award (100 PLN versus 10 PLN). There were four treatment groups: control (with low stake lottery and low expected value), T1 (high stake lottery), T2 (certain 10 PLN (same EV as in the high stake lottery) paid ex post the survey completion), T4 (certain 10 PLN paid ex ante, and unconditional on completing the survey).
Non-financial treatments include different invitation messages: control (no additional info), gaining ("the longer you remain with us, the more valuable the survey is"), losing ("without the help of people like you it would be impossible to create a truly valuable labor market survey in Poland")

Design details

Randomization method
randomization done in office by a computer
Randomization unit
individual subject of the panel online survey DIAGNOZA plus
Clusters
24 304 invitations via email
Observations
24 304 response or non-response (6076 among those who participate at least once)
Arms
financial: control - 5027, T1 - 338, T2 - 376, T3 - 376, T4 - 335 non-financial: control - 2017, gaining - 2088, losing - 1971
Primary outcomes
response rates according to different treatment incentive groups

Cited by

No citation data available for this registration yet.

Keywords